Fusarium graminearum, F. cortaderiae and F. pseudograminearum in New Zealand: molecular phylogenetic analysis, mycotoxin chemotypes and co-existence of species

Mycol. Res. 109 (4): 410–420 (April 2005). f The British Mycological Society

410

doi:10.1017/S0953756204002217 Printed in the United Kingdom.

Fusarium graminearum, F. cortaderiae and F. pseudograminearum in New Zealand : molecular phylogenetic analysis, mycotoxin chemotypes and co-existence of species Russell D. MONDS1#, Matthew G. CROMEY1*, Denis R. LAUREN2, Margaret di MENNA3 and John MARSHALL1 1

New Zealand Institute for Crop & Food Research Limited, Private Bag 4704, Christchurch, New Zealand. The Horticulture and Food Research Institute of New Zealand, Private Bag 3123, Hamilton, New Zealand. 3 AgResearch, Private Bag 3123, Hamilton, New Zealand. E-mail : [email protected] 2

Received 19 June 2004; accepted 1 December 2004.

Fusarium graminearum and F. pseudograminearum are important plant pathogens in New Zealand and around the world. Headblight and crown rot diseases of cereals caused by these species are responsible for large economic losses due to reduction in seed quality and contamination of grain with tricothecene mycotoxins. In the current study we have used two different molecular phylogenetic approaches, AFLPs and gene genealogies, to gain insight into the evolutionary relationships between F. graminearum, and F. pseudograminearum in New Zealand. The worldwide genetic diversity of F. graminearum clade is represented by at least eight biogeographically distinct species (previously designated as lineages of F. graminearum). Our analysis demonstrated that this clade is represented by F. graminearum (=F. graminearum Lineage 7) and F. cortaderiae (=F. graminearum Lineage 8) in New Zealand. Through our analysis we also confirm the presence of F. pseudograminearum in New Zealand as a first record for this organism. Information on species is necessary for preventing the inadvertent intercontinental introduction of genetically unique foreign pathogens associated with world trade. The ability to place species information into a worldwide context enabled postulation that the New Zealand representatives of F. graminearum clade originated from at least two regions, and probably on at least two hosts. Correlation of species descriptions with biogeographical and host information revealed evidence for co-localisation of F. graminearum clade species with potential for genetic outcrossing in the field. Mycotoxin analysis showed F. graminearum (=lineage 7) isolates produce either nivalenol (NIV) or deoxnivalenol (DON). In contrast, F. cortaderiae isolates produced only NIV. These findings support earlier observations that mycotoxin production in the F. graminearum clade is not species specific, but suggest maintenance of chemotype diversity through speciation may have been restricted to a subset of species.

INTRODUCTION In recent years, diseases caused by Fusarium spp. (foot or crown rot and Fusarium Head Blight (FHB)) have become an increasing concern in cereal crops in New Zealand and in other parts of the world (Braithwaite, Alexander & Adams 1998). Epidemics are associated with high economic losses due to contamination of grains with mycotoxins and overall reduction in seed quality (Ichinoe et al. 1983). Various Fusarium species have been linked to FHB, but F. graminearum is emerging as the major causal agent in many parts of the world (O’Donnell et al. 2000). Crown rot can also be caused by a range * Corresponding author. # Current address: Dartmouth Medical School, Department of Microbiology and Immunology, Hanover, NH 03755, USA.

of Fusarium species, but in Australia this disease is most commonly associated with F. pseudograminearum (Aoki & O’Donnell 1999). Fusarium graminearum and F. pseudograminearum are morphologically very similar, and were previously designated Group 1 and Group 2 within F. graminearum (Francis & Burgess 1977). Use of DNA sequence (Aoki & O’Donnell 1999), RFLP (Benyon, Burgess & Sharp 2000) and isozyme (Laday et al. 2000) data indicated that these two fungi were phylogenetically distinct and should be regarded as separate species. Recently, O’Donnell et al. (2000) and Ward et al. (2002) used six gene genealogies to show that the F. graminearum clade (hereafter referred to as the Fg clade) is composed of eight worldwide lineages (one first found in New Zealand). Furthermore, these lineages were biogeographically structured, suggesting

R. D. Monds and others limited gene flow among these populations. The Fg clade lineages have been formally described as phylogenetically distinct species (O’Donnell et al. 2004). In this paper, we use the Fg clade species nomenclature of O’Donnell et al. 2004, with equivalent F. graminearum lineage given at first mention. FHB is common in wheat in New Zealand, especially in North Island crops where the disease is primarily associated with F. graminearum (Sayer & Lauren 1991, Cromey et al. 2002). The disease is less common in South Island crops, where it is most often associated with F. culmorum and F. avenaceum (Cromey, Parkes & Fraser 2001). Braithwaite et al. (1998) commonly observed F. avenaceum and F. graminearum associated with foot rot of cereals. However, F. graminearum isolates were not classified as Groups 1 and 2, and may have been F. pseudograminearum, although this species had not yet been recorded in New Zealand. Strains of both F. graminearum and F. pseudograminearum have been shown to produce trichothecene mycotoxins (Abramson, Clear & Smith 1993, O’Donnell et al. 2000). Three major strain-specific chemotypes have been identified within the B-trichothecene lineage of with Fusarium : nivalenol (NIV chemotype), deoxynivalenol and 3-acetyldeoxynivalenol (3ADON chemotype), and deoxynivalenol and 15-acetyldeoxynivalenol (15ADON chemotype) (Ward et al. 2002). DON chemotypes are most commonly associated with F. graminearum, although some isolates produce NIV (Abramson et al. 1993, O’Donnell et al. 2000, Gilbert et al. 2001). In New Zealand, however, Lauren et al. (1992) found that NIV chemotypes were as common as DON chemotypes. Less information is available on mycotoxin production by strains of F. pseudograminearum, probably because this species is not normally associated with cereal grains used for food or feed. However, Ward et al. (2002) found that three isolates of F. pseudograminearum produced 3ADON, while a fourth isolate was of unknown chemotype. Information on Fg clade species in New Zealand could also provide information on the origin of the New Zealand population, as well as potential pathways of entry for fungal plant pathogens. Such knowledge will assist in preventing the inadvertent intercontinental introduction of genetically unique foreign populations of pathogenic organisms associated with world trade (O’Donnell et al. 2000). Outcrossing between Fg clade species has resulted in novel pathogens (O’Donnell et al. 2000). Further hybridisation between Fg clade species is possible, given the wide host range of this clade. In this study we examined 34 New Zealand isolates of F. graminearum/F. pseudograminearum (the two species were not differentiated prior to the commencement of the study) from wheat, maize and barley grains and stem bases. Seven F. culmorum isolates from stem bases or grains were also examined. Phylogenies were constructed using both AFLP and gene genealogy strategies to determine the evolutionary relationships

411 of these species in New Zealand. Correlation of phylogenetic data with that of O’Donnell et al. (2000) has allowed us to place our results within a worldwide context, offering insight into the origin and entry of F. graminearum isolates into New Zealand. Speciesmycotoxin relationships were also investigated providing general support for maintenance of chemotype polymorphisms in a subset of Fg clade species.

MATERIALS AND METHODS Media and culturing conditions Fusarium species from wheat, barley or maize were isolated by surface-sterilising cereal grains or stem bases in 1 % sodium hypochlorite for 10 min, rinsing in sterile water, and plating onto Potato Dextrose agar (PDA) amended with antibiotics (120 mg mlx1 streptomycin, 50 mg mlx1 aureomycin and 50 mg mlx1 chloramphenicol). Plates were incubated for 4–8 d at 25 xC under conditions of alternating light and dark (12 h photoperiod). Single spore isolates of selected Fusarium colonies were obtained using the method of Burgess et al. (1994). Morphological identification Species identification was carried out after culturing isolates on PDA and carnation-leaf agar (CLA) using the methods and criteria of Burgess et al. (1994). Plates for species identification were incubated for 7 days in an alternating temperature regime, 25 x light/20 x dark, with a 12 h photoperiod. Colony diameters were assessed for selected isolates on PDA after dark incubation for 3 d at both 25 x and 30 x. Perithecium production Isolates were screened for perithecium production on hay agar (2 g chopped meadow hay, 15 g agar, 1 l distilled water). Culture plates were incubated on a light bench (40 W cool white fluorescent tubes mounted 25 cm above bench height, 12 h photoperiod, no temperature regulation) for up to 10 wk with regular examination for development of perithecia. Isolates used for molecular analysis A range of isolates identified as Fusarium graminearum, F. pseudograminearum, or F. culmorum using morphological characters were selected for molecular analysis (Table 1). They included isolates from different regions (the North or South Islands of New Zealand), different hosts (maize, wheat or barley) and different host tissues (grains or stem bases). Maize isolates were from two samples of maize kernels of hybrid 36R10 ; harvested on 10 May 2000 from a Te Awamutu (North Island) property, or harvested on 20 June 2000 from a Taumaranui (North Island) property. Cultures of each

Fusarium graminearum and F. pseudograminearum

412

Table 1 Fungal strains used in this study, their source, and their perithecial production, growth rate, and mycotoxin production.

ICMP accession no.

Growth rate 25 xC Host

Tissue

Locationa

Perithecium production

F. graminearum NZ62 15504 NZ61 15503 NZ66 15507 NZ53 15495 NZ54 15496 NZ52 15494 NZ65 15506 NZ64 15505 NZ33 15483 NZ31 15482 NZ12 15481 NZ35 15484 NZ51 15493 NZ49 15491 NZ70 15508

Maize Maize Wheat Maize Maize Maize Wheat Wheat Wheat Wheat Wheat Wheat Maize Maize Wheat

Grain Grain Grain Grain Grain Grain Grain Grain Grain Grain Grain Grain Grain Grain Healthy stem

Te Awamutu, NI Te Awamutu, NI Kairanga, NI Te Awamutu, NI Te Awamutu, NI Te Awamutu, NI Marton, NI Marton, NI Kairanga, NI Kairanga, NI Kairanga, NI Kairanga, NI Te Awamutu, NI Te Awamutu, NI Marton, NI

+ + + + + + + + + + + + + + +

F. cortaderiae NZ43 15485 NZ44 15486 NZ45 15487 NZ46 15488 NZ47 15489 NZ48 15490 NZ50 15492 NZ55 15497 NZ56 15498 NZ57 15499 NZ58 15500 NZ59 15501 NZ60 15502

Maize Maize Maize Maize Maize Maize Maize Maize Maize Maize Maize Maize Maize

Grain Grain Grain Grain Grain Grain Grain Grain Grain Grain Grain Grain Grain

Taumaruanui, NI Taumaruanui, NI Taumaruanui, NI Taumaruanui, NI Taumaruanui, NI Taumaruanui, NI Te Awamutu, NI Taumaruanui, NI Taumaruanui, NI Taumaruanui, NI Taumaruanui, NI Te Awamutu, NI Te Awamutu, NI

F. culmorum NZ6 15476 NZ3 15474 NZ4 15475 NZ7 15477 NZ34 15480 NZ23 15478 NZ24 15479

Wheat Wheat Barley Wheat Wheat Wheat Wheat

Healthy stem Stem lesion Healthy stem Healthy stem Grain Healthy stem Stem lesion

Dorie, SI Temuka, SI Dunearn, SI Marton, NI Kairanga, NI Irwell, SI Eiffelton, SI

F. pseudograminearum NZ69 15513 NZ68 15512 NZ5 15509 NZ67 15514 NZ38 15511 NZ32 15510

Wheat Wheat Barley Wheat Wheat Wheat

Stem lesion Stem lesion Healthy stem Grain Grain Stem lesion

Morven, SI Darfield, SI St Andrews, SI Kairanga, NI Kairanga, NI Marton, NI

Isolate

a

Growth rate 30 xC

mm dx1

9.2

7.7

11.0

7.5

10.8

7.7

10.8

7.7

10.8

7.3

+ + + + + + + + + + + + +

11.2 10.5

6.0 6.2

9.3

3.3

10.8 12.0

6.7 7.3

x x x x x x

8.8 10.0 10.2 8.7 8.8

2.3 4.8 3.5 6.3 6.3

DON NIV

3ADON

15ADON

+++ n.d. n.d. n.d. +++ +++ n.d. n.d. +++ n.d. +++ n.d. +++ +++ n.d.

n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d.

+ n.d. +++ +++ n.d. n.d. +++ n.d. + +++ + +++ n.d. + +++

+++ +++ +++ +++ +++ +++ +++ ++ +++ ++ +++ +++ +++

n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d.

+ + n.d. n.d. n.d. n.d. n.d. n.d. + n.d. + + +

n.d. + + +++ +++ ++

+++ +++ +++ n.d. n.d. n.d.

n.d. n.d. n.d. + + n.d.

NI, North Island; and SI, South Island.

isolate have been deposited in the ICMP collection of fungus cultures (Landcare Research, Private Bag 92170, Auckland, NZ ; Table 1). DNA preparation Single spore isolates were used to inoculate 25 ml of Potato Dextrose Broth (PDB) which was then incubated for 7 d at 22 x in an orbital shaker. Mycelium was harvested from liquid cultures by filtration onto filter paper (Whatman grade 114). After rinsing with sterile water, the mycelium was frozen in liquid nitrogen

and ground to a fine powder with a mortar and pestle. A slurry was formed by the addition, and mixing, of 10 ml CTAB buffer (2 % hexadecyltriethylammonium bromide, 1.4 M NaCl, 20 mM Tris pH 8.0, 20 mM EDTA pH 8.0, 0.001 % mercaptoethanol), which was then transferred to a 50 ml centrifuge tube and incubated at 65 x for 1.5 h. An equal volume of chloroform : iso-amyl alcohol (24 :1) was added, mixed and centrifuged at 3500 rpm for 15 min. The aqueous phase was transferred to a 50 ml falcon tube and total DNA precipitated by the addition of one volume of 100 % ethanol. After incubation at room temperature for

R. D. Monds and others 10 min, DNA was spooled from solution and placed in wash solution I (76 % ethanol, 0.2 M sodium acetate) for 20 min at room temperature. Subsequently, spooled DNA was placed in wash solution II (76 % ethanol, 0.01 M ammonium acetate) for 1 min before being placed in an Eppendorf tube and air dried. DNA was dissolved in 300–400 ml of 10 mM Tris HCl, 1 mM EDTA pH 8.0 at 4 x overnight before further manipulation. AFLP analyses AFLP methodology was essentially as described by Barrett & Kidwell (1998) and Pickering et al. (1995) with minor modifications. Templates were prepared by digesting 500 ng of sample DNA with PstI and MseI for 3 h. Pst and Mse adaptors (Pickering et al. 1995) were then ligated to restricted genomic fragments for another 3 h at 37 x. Two selective rounds of PCR were performed : a preamplification and a 32P labelled PCR. Pre-amplification was carried out using Mse+1 and Pst+1 primers ; either Mse+G/Pst+A or Mse+G/Pst+G primer pairs. For the 32P labeled PCR, a combination of Mse+3 primers in concert with Pst+1 primers were used. Samples were electrophoresed in a denaturing 6 % 29 :1 acrylamide/bis-acrylamide gel for 2 h and 10 min at 70 W. The gels were then dried on to filter paper (Whatman 3MM). For visualisation, dried gels were exposed to autoradiographic film for 1–2 d before developing using standard procedures. DNA Sequencing Phosphate permase (Pho) sequences were amplified from total DNA with primers PHO1 and PHO6 (O’Donnell et al. 2000). Reactions contained 0.2 mM of each primer, 40 mM dNTP’s and 0.6 units Taq DNA polymerase (Roche) in a final volume of 25 ml. Thermal cycling conditions were 40 times, 94 x for 30 s; 50 x for 30 s; 72 x for 1 min. PCR products were sequenced directly using Big Dye v3.0 (Applied Biosystems, Foster City, CA) and primer PHO1. Sequences were collated and edited using SEQUENCHER TM 3.0 sequence analysis software package (Gene Codes Corporation, Ann Arbor, MI). Phosphate permase sequences obtained from external sources were treated in a similar manner to allow comparative analyses. Phylogenetic analysis AFLP data. AFLP bands were scored as either 1 (presence of a band) or 0 (absence of a band). In total, 155 polymorphic bands were scored for the 41 isolates used in this study. Scores were used to generate a binary matrix for all isolates and phylograms constructed using the software package TREECON for Windows

413 version 1.3b (van de Peer & Wachter 1994). The Nei & Li method was used to estimate genetic distance (Nei & Li 1979) and neighbour-joining methods were used to infer tree topology (Saitou & Nei 1987). Relative support for individual nodes was obtained with bootstrap analysis using 1000 pseudo-replicates (Felsenstein 1985). The tree was rooted to the F. pseudograminearum clade. Sequence data. Pho sequences were aligned using CLUSTALW (Thompson, Higgins & Gibson 1994) and a phylogram constructed from these alignments using TREECON (van de Peer & Wachter 1994). The Jukes and Cantor model was used to estimate genetic distance between sequences (Jukes & Cantor 1969), and tree topology was inferred using neighbor joining methods (Saitou & Nei 1987). Support for individual nodes was obtain by calculating bootstrap values based on the analysis of 1000 pseudo-replicates (Felsenstein 1985). The tree was rooted to the F. pseudograminearum clade. Mycotoxin analyses Discs (1 cm diam) from 4 d old Fusarium cultures on PDA were used to inoculate 50 ml of MYRO liquid medium, (NH4)2HPO4 (1 g), KH2PO4 (3 g), MgSO4 .7H2O (0.2 g), NaCl (5 g), sucrose (40 g), glycerol (10 g), and ultra-pure water to 1 l (Lauren, Sayer & di Menna 1992), in 250 ml conical flasks. The flasks were incubated on a shaker at 23 x for 14 d. The total cultures were extracted with ethyl acetate (2r50 ml) and the collected extract evaporated to dryness. An aliquot of the dried residue was hydrolysed to produce parent alcohols prior to analysis by HPLC with UV detection at both 245 nm and 254 nm for NIV and DON. The detection limit was 0.2 mg flaskx1 for NIV and DON. Further aliquots of selected samples which produced DON were analysed without the hydrolysis step and using LC-MS (liquid chromatography – mass spectrometry) to determine whether the dominant form of DON-chemotype was 3ADON or 15ADON.

RESULTS Isolation of Fusarium species A range of Fusarium spp. were isolated from cereal grains and stem bases (Table 1). These included F. culmorum, F. avenaceum, F. crookwellense, F. poae, F. oxysporum, F. heterosporum, F. equiseti as well as F. graminearum/F. pseudograminearum. All maize isolates were from grains, while isolates from wheat and barley were from stem bases as well as grains. An attempt was made to separate isolates based on measuring the region of maximum width of conidia. However, variability within isolates was too great for differences between isolates to be meaningful. Single spore cultures of the 34 isolates identified as F. graminearum/F. pseudograminearum were tested

Fusarium graminearum and F. pseudograminearum for perithecial production. Of these, 28 isolates of the Fg clade produced perithecia, while the six isolates of F. pseudograminearum did not (Table 1). All 20 isolates from maize grains were perithecium producers, while seven of nine isolates from wheat grains and an isolate from a wheat stem produced perithecia. These 28 isolates were all from the North Island of New Zealand, and were putatively assigned to the homothallic species F. graminearum (previously group 2). Where 3-septate ascospores were measurable they were 20–24 mm long, agreeing with the measurement for Gibberella zeae ascospores given by Aoki & O’Donnell (1999). Six isolates identified as F. graminearum/ F. pseudograminearum did not produce perithecia, and were putatively assigned to the heterothallic species F. pseudograminearum. Three of these isolates were from the North Island, two being isolated from wheat grains and one from a wheat stem. The other three isolates were from the South Island and were isolated from wheat or barley stems. The mean growth rates on PDA at 25 x or 30 x differed for the perithecium-producers and nonproducers. The mean of ten perithecium producers at 25 x was 10.7 mm dayx1 compared with a mean for five isolates without perithecia of 9.3 mm dayx1. The difference at 30 x was greater (6.7 mm dayx1 and 4.7 mm dayx1 respectively). At 25 x, there was some overlap amongst the individual isolates. At 30 x there was also some overlap, with one peritheciumproducing isolate having a particularly slow growth rate. The 28 Fg clade isolates, six F. pseudograminearum isolates and seven F. culmorum isolates from New Zealand wheat or barley stems or grains, were selected for molecular phylogenetic analysis. AFLP analyses Isolates of the Fg clade, Fusarium pseudograminearum and F. culmorum were subjected to phylogenetic analysis using AFLPs. Eight separate Pst+1/ Mse+3 primer combinations were used to generate unique amplification profiles, from which 155 polymorphic markers were scored for their presence or absence. Although showing conserved banding patterns, profiles for each species were qualitatively distinguishable by eye. The phylogram constructed from AFLP data (Fig. 1) supported the species groupings assigned on the basis of morphology and qualitative observation of AFLP profiles. Division of F. pseudograminearum and Fg clade species into monophyletic groups was supported by high bootstrap values. In contrast, although all F. culmorum isolates formed a monophyletic group, bootstrap analysis only provided moderate relative support for this division. Fg clade isolates formed two reciprocally monophyletic groups, suggesting that significant genotypic differences exist within the New Zealand population of F. graminearum. Within the context of AFLP analysis

414 we refer to these lineages as Group A (later confirmed as F. graminearum,=F. graminearum lineage 7) and B (later confirmed as F. cortaderiae (=F. gramineraum lineage 8)). In the case of group B (F. cortaderiae) this division was supported by a bootstrap value of 100%. For Group A (F. graminearum), a bootstrap value of 69 % was obtained in support of the monophyletic assignment. Although this value is much lower than for F. cortaderiae isolates, on viewing the tree topology, it seems likely that the long branch length of NZ70 is largely responsible for lowering the statistical confidence in designating F. graminearum as monophyletic. NZ70 is considered an outlier in F. graminearum and differs from other F. graminearum members’ tissue origin, being isolated from stem tissue rather than grain. F. pseudograminearum isolates formed two reciprocally monophyletic lineages. In this instance both groupings were statistically supported by bootstrap values over 97 %. This genotypic division was associated with geographical division into North and South Island isolates. Furthermore, branching of South Island F. pseudograminearum isolates could be correlated with differences in host origin (wheat or barley, Fig. 1), and also demonstrated bootstrap values over 90 %. Gene genealogy analysis To further substantiate the existence of two Fg clade species in New Zealand, as well as to place our results in a worldwide context, we adopted the same gene genealogy strategy as that utilised by O’Donnell et al. (2000). Phosphate permase (Pho) sequence from isolates in this study were combined with the Pho sequence data for Fg clade, Fusarium culmorum and F. pseudograminearum isolates from O’Donnell et al. (2000). A phylogenetic analysis was performed for all 76 isolates using 644 bp of aligned Pho sequence (Fig. 2). As with the AFLP-based phylogram, the sequence based phylogeny supported morphological separation of Fg clade, F. culmorum and F. pseudograminearum. The seven Fg clade species described by O’Donnell et al. (2000, 2004) were conserved in the Pho analysis. Concordant with the AFLP phylogeny, F. graminearum and F. cortaderiae isolates were once again resolved as phylogenetically distinct. Using the Pho sequence, Gram 47 clustered with F. austroamericanum (=F. graminearum Lineage 1) isolates, not F. cortaderiae as seen in the AFLP based phylogram. The six F. pseudograminearum isolates grouped with the four isolates used in the study of O’Donnell et al. (2000), however the phylogram suggested the existence of three F. pseudograminearum lineages. The isolates of O’Donnell et al. (2000) formed one distinct lineage, and two lineages of New Zealand isolates were also resolved. Tree topology for New Zealand isolates differed from that derived from AFLPs in that isolate NZ67, from a North Island wheat grain, joined isolates

R. D. Monds and others

415

10 % Genetic distance NZ62 NI maize grain

76

NZ61 NI maize grain NZ66 NI wheat grain NZ53 NI maize grain NZ54 NI maize grain

F. graminearum

NZ52 NI maize grain NZ65 NI wheat grain

98

NZ64 NI wheat grain NZ33 NI wheat grain NZ31 NI wheat grain

94

62

NZ12 NI wheat grain

73

NZ35 NI wheat grain

74

NZ70 NI wheat stem 69

NZ51 NI maize grain NZ49 NI maize grain 88

96

NZ50 NI maize grain NZ48 NI maize grain NZ60 NI maize grain

100

NZ58 NI maize grain

NZ56 NI maize grain NZ59 NI maize grain NZ55 NI maize grain

100

F. cortaderiae

NZ57 NI maize grain

66

NZ47 NI maize grain NZ45 NI maize grain NZ43 NI maize grain NZ46 NI maize grain NZ44 NI maize grain NZ6 SI wheat stem

NZ4 SI barley stem NZ3 SI wheat stem NZ24 SI wheat stem

F. culmorum

NZ23 SI wheat stem

65

NZ34 NI wheat grain

100

NZ69 SI wheat stem

91

NZ68 SI wheat stem

98

NZ5 SI barley stem

100

NZ67 NI wheat grain 100 94

NZ38 NI wheat grain NZ32 NI wheat stem

F. pseudograminearum

NZ7 NI wheat stem

Fig. 1. Phylogram of total AFLP results for Fusarium graminearum, F. cortaderiae, F. pseudograminearum and F. culmorum. Tree topology was inferred using Neighbour Joining methods, bootstrap values >55 % are shown above nodes and were calculated from 1000 pseudo-replicates. Origins of isolates are indicated after the isolate number (NI, North Island ; and SI, South Island).

Fusarium graminearum and F. pseudograminearum

416

2% Genetic distance NZ66 wheat NZ12 wheat 28336 NZ62 maize NZ31 wheat NZ70 wheat NZ61 maize NZ49 maize NZ52 maize NZ35 maize 5883 NZ53 maize NZ54 maize NZ51 maize NZ65 wheat NZ64 wheat 13383 29169 28063 6394 28439 26156 13818 63 28721 28720 6101 96 26755 26754 26752

57

100

79 96

92

F. graminearum

26916 29020 29011 29105 95 29148 25797 28723 28436 29010 NZ47 maize 2903 60 28585 28718 84 29297_2 62 63 NZ57 maize NZ55 maize NZ60 maize NZ56 maize NZ45 maize NZ50 maize 29306 NZ44 maize NZ43 maize NZ48 maize NZ46 maize NZ59 maize NZ58 maize 65

100

F. austroamericanum

F. cortaderiae

56

25475 3288 NZ7 wheat NZ6 wheat

F. culmorum

70

NZ4 barley NZ3 wheat NZ23 wheat NZ34 wheat 100 NZ24 wheat

100 99

28065 28062 28334 28338

NZ38 wheat NZ32 wheat NZ68 wheat NZ69 wheat NZ67 wheat NZ5 barley 100

62 100 78

F. pseudograminearum

Fig. 2. Phylogram of Fusarium isolates from this study (isolate number and host crop) and O’Donnell et al. (2000) generated using sequence from the phosphate permase gene. The Jukes and Cantor model was used to estimate genetic distance, tree topology was inferred using Neighbour Joining methods, and bootstrap values >55 % are shown above nodes calculated from 1000 pseudoreplicates. Relevant Fg clade species as described by O’Donnell et al. (2004) are indicated.

R. D. Monds and others 5, 68, and 69, which were from South Island cereal stems. For both analylses statistical support for respective monophyletic groupings was high, however it should be noted that AFLP based analyses sample heterogeneity over the entire genome whereas the Pho sequence analysis samples one locus.

417 isolates, three produced DON (all 3ADON) while the other three produced NIV (two of them also producing small quantities of 15ADON). The three DON producers were all isolated from South Island cereal stems, while the three NIV producers were from North Island cereals (two from grains and one from a stem lesion).

Potential for gene flow amongst NZ Fg clade species Species designations for each Fg clade isolate were correlated with available information on the region of isolation and host origin as a means to assess the potential for gene flow between species (Table 1, Fig. 1). Within our data set we identified one instance of geographical as well as host co-localisation. Both Fusarium graminearum and F. cortaderiae were isolated from maize grains at the North Island (N.I.) Te Awamutu site. Out of ten isolates, seven were F. graminearum and three were F. cortaderiae. Conversely, all ten isolates derived from maize ex grain at the Tauramanui site were F. cortaderiae. Isolates derived from wheat were also shown to be homogeneous, consisting solely of F. graminearum. In summary, these correlations show that at least two Fg clade species can co-exist in similar niches making gene flow a formal possibility. The isolation of only one species from wheat and from the Tauramanui site is harder to interpret. This data may suggest that, although colocalisation does occur, specific niche presences exist for the two species. However, a larger sampling size may well demonstrate co-localisation as was seen at the Te Awamutu site. Mycotoxin analyses The 15 isolates of Fusarium graminearum, the 13 isolates of F. cortaderiae, and the six isolates of F. pseudograminearum were assayed for their production of the trichothecene mycotoxins NIV and/or DON (either 3ADON or 15ADON). All F. cortaderiae isolates were NIV producers (Table 1), which agrees with the mycotoxin production by two isolates of this species analysed by Ward et al. (2002). Six of the isolates also produced small quantities of 15ADON, while none produced detectable levels of 3ADON. Seven of the F. graminearum isolates were NIV producers (four of these also produced small quantities of 15ADON), while six were DON (all 15ADON) producers. Two F. graminearum isolates did not produce detectable levels of either mycotoxin. When the results of F. graminearum mycotoxin analyses are viewed in the context of host origin, both chemotypes were associated with maize and wheat. Specifically, of the F. graminearum maize isolates, five were NIV producers, one was a DON producer, whereas of the eight F. graminerum isolates from wheat, two were NIV producers and five were DON producers. These results suggest a lack of mycotoxin correlation with F. graminearum isolates and their host origin. Of the six F. pseudograminearum

DISCUSSION In this study we utilised two different molecular phylogenetic approaches, AFLPs and gene genealogies, to gain insight into the evolutionary relationships between members of the Fg clade and Fusarium pseudograminearum in New Zealand. Furthermore, we place our results in a worldwide context by adopting the same framework for gene genealogy analysis as described by O’Donnell et al. (2000). Analyses of mycotoxin chemotypes and subsequent correlation with biogeographic, host and species information is discussed in the context of between-species gene flow and biosecurity. Phylogenetic analysis was carried out on 28 Fg clade isolates, six F. pseudograminearum isolates and seven F. culmorum isolates which were assigned to species groupings on the basis of morphological and physiological criteria. Phylogenies reconstructed both from AFLP and phosphate permase sequence data indicate the presence of two distinct Fg clade species, F. graminearum and F. cortaderiae (O’Donnell et al. 2004), in New Zealand. Members of the Fg clade occur worldwide, and comprise nine biogeographically structured phylogenetically distinct species, indicating a long evolutionary history of reproductive isolation (O’Donnell et al. 2004). The ability to place the New Zealand species of the Fg clade within the context of the worldwide phylogeny provides avenues to gaining insight into the origin and possible route of entry of these pathogens into New Zealand. A recently identified species (F. cortaderiae), although first detected in New Zealand, is thought to be of South American origin (O’Donnell et al. 2004). This species was isolated from pampas grass (Cortaderia jubata), suggesting it may have arrived in New Zealand on imported plants. Pampas grass occurs widely in New Zealand as a shelter and ornamental plant and as a weed, particularly in forestry areas, but originates from South America. The airborne ascospores of F. cortaderiae could assist in the spread of this species from hosts such as pampas grass to cereal crops. F. graminearum, the other Fg clade species found in the present study, is presently cosmopolitan in its distribution (O’Donnell et al. 2004). Cereal (wheat, barley, maize) imports to New Zealand are primarily from Europe and North America, as well as Australia. O’Donnell et al. (2000) recorded F. graminearum

Fusarium graminearum and F. pseudograminearum commonly from Europe and USA. O’Donnell (pers. comm.) had previously only detected this species from cereals in New Zealand. It is likely, then, that it arrived on cereal grain imports to New Zealand. One isolate (NZ47) clustered with F. austroamericanum using the Pho sequence, but with F. cortaderiae in the AFLP based phylogram. However, there was no indication that this isolate was an outlier in the AFLP analysis. Further gene genealogies will need to be constructed to discern between the possibilities that NZ47 is either more accurately grouped with F. austroamericanum, or is the result of hybridisation. Otherwise, two distinct analyses gave the same two clear groups of isolates, corresponding to F. graminearum and F. cortaderiae. There was no evidence for consistent groups of isolates below the level of species; subspecific grouping of isolates differed between the two analyses. In the past, fungi have been considered recalcitrant to inter or intra-specific genetic exchange. However, the last decade has seen increasing numbers of reports documenting specific hybridisation events amongst natural populations (Olsen & Stenlid 2002). The identification of a hybrid strain between two Fg clade species (O’Donnell et al. 2000) and demonstration of laboratory outcrossing amongst lineages (Bowden & Leslie 1999) indicates that novel lineages could develop given the appropriate geographical co-localisation. In the present study we have documented co-localisation of F. graminearum and F. cortaderiae in a New Zealand maize crop. Out of the ten isolates from the Te Awamutu site, seven were F. graminearum and three were F. cortaderiae. Although this is a single instance, such a finding indicates the potential for co-localisation of these species as well as the potential for genetic outcrossing. No clear hybrids between these species were detected in the present study. Given the percentage of the genome sampled, however, this is not unexpected. An extended phylogenetic analysis of the co-localised species isolated in this study may uncover evidence for hybridisation. Presence of perithecia, molecular phylogenetic analysis, mycotoxin chemotype and, to a lesser extent, growth rate and host tissue associations indicate that two Fg clade species and F. pseudograminearum exist on cereals in New Zealand. To our knowledge this is the first clear documentation of F. pseudograminearum in New Zealand. Francis & Burgess (1977) and Burgess et al. (1994) suggest that F. graminearum and F. psuedograminearum cannot be reliably differentiated on the basis of morphology of macroconidia. We were also unable to confidently discern consistent conidial shape differences between isolates of F. graminearum and F. pseudograminearum, although Aoki & O’Donnell (1999) found small differences in conidia of the two species using different media and culture conditions. Presence or absence of perithecia was therefore used to initially classify isolates into the two groups. On average, the putative Fg clade isolates grew faster

418 than the putative F. pseudograminearum isolates, as previously recorded by Burgess et al. (1994) and Aoki & O’Donnell (1999). However, there was considerable overlap between individual isolates. Therefore, growth rate was not a reliable indicator of species in itself, although it did support the species separation based on perithecia. F. pseudograminearum is most commonly associated with crown rot of cereals, while F. graminearum is most commonly associated with head blight (Francis & Burgess 1977). Likewise, in the present study, almost all the isolates assigned to the Fg clade came from cereal grains, while most isolates assigned to F. pseudograminearum were from cereal stems. Phylogenetic analysis confirmed the separation of the two groups of isolates, and comparison with Pho permease sequence data for known isolates of the Fg clade and F. pseudograminearum indicated that the putative designations were appropriate (O’Donnell et al. 2000). Analysis of a limited number of isolates of F. pseudograminearum suggests some phylogenetic diversity within this species. Previous reports suggest that F. pseudograminearum isolates produce the mycotoxin 3ADON rather than NIV (O’Donnell et al. 2000). Of six isolates in the present study assigned to F. pseudograminearum, three were 3ADON producers and three were NIV producers. The three 3ADON producing isolates were obtained from cereal stems in South Island crops, and formed one of the two clusters of isolates indicated by AFLP analysis. These isolates would appear to be classic F. pseudograminearum isolates based on mycotoxin chemotype and host tissue association. The three NIV producers (two of which also produced small quantities of 15ADON) formed the other cluster, and were isolated from North Island cereal crops, two isolates being from grains and one from a stem lesion. The AFLP data therefore suggest two distinct populations separated by geography and mycotoxin chemotype. However, using the phosphate permase gene sequence, one of the North Island isolates (isolated from a grain) clustered with the South Island isolates. Morphological comparison of New Zealand F. pseudograminearum isolates with typical strains has confirmed their identification (Alison Bentley, pers. comm.). Further work with a greater number of isolates is being carried out to determine the population structure of F. pseudograminearum in New Zealand, and whether New Zealand isolates may include additional undescribed species closely related to F. pseudograminearum (A. Bentley, unpublished). Biosecurity (plant quarantine) has become an increasingly important consideration throughout the world. Reflecting this concern, Ward et al. (2002) suggest that ‘disease control and plant quarantine programmes need to exercise increased vigilance to ensure that international trade in agricultural commodities does not result in the global transposition of strains with novel toxigenic potential ’. Indeed,

R. D. Monds and others Benyon et al. (2000) using RFLP and Ouellet & Seifert (1993) using RAPD-PCR found that isolates of F. graminearum could not be grouped according to host or geographic origin. This finding suggests that these genetically similar organisms may have been dispersed from a common point of origin by the airborne dissemination of spores over time, or more recently by the movement of seed stock or contaminated agricultural implements. In this study we show the presence of two distinct Fg clade species in New Zealand which are not likely to face substantial geographical/host barriers to hybridisation. Conversely, we did not detect several Asian or New World Fg clade endemics. Cereal grain imports to New Zealand from Asia, South America, and Africa are much more closely controlled by quarantine measures (NZ Seed for Sowing Standard) than imports from Australia, North America and Europe. This may have prevented entry to New Zealand of the common species from Asia and South America. Because some of the Fg clade species are reported to have a wide host range, including non-cereals, it will be important to characterise New Zealand isolates from a range of hosts. Hybrids formed via genetic outcrossing of introduced species with resident species could have novel host range, mycotoxin and pathogenicity characteristics, and quarantine measures to limit the entry or spread of further species is warranted.

ACKNOWLEDGEMENTS Rochelle Parkes, Wendy Smith, Dwayne Jensen, Tania Trower and Claire Nichols provided technical assistance. We thank Gail Timmerman-Vaughan and Tonya Frew for assistance with AFLPs. Kerry O’Donnell kindly supplied Pho sequence data and provided valuable comments on the manuscript. We also thank Ryan Thum and Scott Godfrey for comments on the manuscript. Funding for this research was provided by the New Zealand Foundation for Research Science and Technology.

REFERENCES Abramson, D., Clear, R. M. & Smith, D. M. (1993) Trichothecene production by Fusarium spp. isolated from Manitoba grain. Canadian Journal of Plant Pathology 15: 147–152. Aoki, T. & O’Donnell, K. (1999) Morphological and molecular characterisation of Fusarium pseudograminearum sp. nov, formerly recognised as the Group 1 population of F. graminearum. Mycologia 91 : 597–609. Barrett, B. A. & Kidwell, K. K. (1998) AFLP-based genetic diversity assessment among wheat cultivars from the Pacific Northwest. Crop Science 38 : 1261–1271. Benyon, F. H. L., Burgess, L. W. & Sharp, P. J. (2000) Molecular genetic investigations and reclassification of Fusarium species in sections Fusarium and Roseum. Mycological Research 104: 1164–1174. Bowden, R. L. & Leslie, J. F. (1999) Sexual recombination in Gibberella zeae. Phytopathology 89: 182–188. Braithwaite, M., Alexander, B. J. R. & Adams, R. L. M. (1998) Nationwide survey of pests and diseases of cereal and grass seed

419 crops in New Zealand. 2. Fungi and Bacteria. Proceedings of the 51st New Zealand Plant Protection Conference : 51–59. Burgess, L. W., Summerell, B. A., Bullock, S., Gott, K. P. & Backhouse, D. (1994) Laboratory Manual for Fusarium Research. 3rd edn. University of Sydney, Sydney. Cromey, M. G., Parkes, R. A. & Fraser, P. M. (2001) Fusarium levels in grain harvested from New Zealand wheat and barley crops in 2000. New Zealand Plant Protection 54: 193–197. Cromey, M. G., Shorter, S. C., Lauren, D. R. & Sinclair, K. I. (2002) Cultivar and crop management influences on Fusarium head blight and mycotoxins in spring wheat (Triticum aestivum) in New Zealand. New Zealand Journal of Crop and Horticultural Science 30: 235–247. Felsenstein, J. (1985) Confidence limits on phylogenies: an approach using the bootstrap. Evolution 39: 783–791. Francis, R. G. & Burgess, L. W. (1977) Characteristics of two populations of Fusarium roseum ‘Graminearum’ in eastern Australia. Transactions of the British Mycological Society 68: 421–427. Gilbert, J., Abramson, D., McCallum, B. & Clear, R. (2001) Comparison of Canadian Fusarium graminearum isolates for aggressiveness, vegetative compatibility, and production of ergosterol and mycotoxins. Mycopathologia 153: 209–215. Ichinoe, M., Kurata, H., Sugiura, Y. & Ueno, Y. (1983) Chemotaxonomy of Gibberella zeae with special reference to production of trichothecenes and zealenone. Applied Environmental Microbiology 46 : 1364–1369. Jukes, T. H. & Cantor, C. R. (1969) Evolution of protien molecules. In Mammalian Protein Metabolism (H. N. Munro, ed.): 21–132. Academic Press, New York. Laday, M., Bagi, F., Mesterhazy, A. & Szecsi, A. (2000) Isozyme evidence for two groups of Fusarium graminearum. Mycological Research 104: 788–793. Lauren, D. R., Sayer, S. T. & di Menna, M. E. (1992) Trichothecene production by Fusarium species isolated from grain and pasture througout New Zealand. Mycopathologia 120: 167–176. Nei, M. & Li, W. H. (1979) Mathematical model for studying genetic variation in terms of restriction endonucleases. Proceedings of the National Academy of Sciences, USA 76: 5269– 5273. O’Donnell, K., Kistler, H. C., Tacke, B. K. & Casper, H. H. (2000) Gene genealogies reveal global phylogeographic structure and reproductive isolation among lineages of Fusarium graminearum, the fungus causing wheat scab. Proceedings of the National Academy of Sciences, USA 97: 7905–7910. O’Donnell, K. O., Ward, T. J., Geiser, D. M., Kistler, H. C. & Aoki, T. (2004) Genealogical concordance between the mating type locus and seven other nuclear genes supports formal recognition of nine phylogentically distinct species within the Fusarium graminearum clade. Fungal Genetics and Biology 41: 600–623. Olsen, A. & Stenlid, J. (2002) Pathogenic fungal species hybrids infecting plants. Microbes and Infection 4: 1353–1359. Pickering, R. A., Hill, A. M., Michel, M. & TimmermanVaughan, G. M. (1995) The transfer of a powdery mildew resistance gene from Hordeum bulbosum L. to barley (H. vulgare L.) chromosome 2 (2L). Theoretical and Applied Genetics 91: 1288– 1292. Ouellet, T. & Seifert, K. A. (1993) Genetic characterisation of Fusarium graminearum strains using RAPD and PCR amplification. Phytopathology 83: 1003–1007. Saitou, N. & Nei, M. (1987) The neighbor-joining method: a new method for reconstructing phylogenetic trees. Molecular Biology and Evolution 4: 406–425. Sayer, S. T. & Lauren, D. R. (1991) Fusarium infection in New Zealand grain. New Zealand Journal of Crop and Horticultural Science 19: 143–148. Thompson, J. D., Higgins, D. G. & Gibson, T. J. (1994) CLUSTAL W: improving the sensitivity of progressive multiple sequence

Fusarium graminearum and F. pseudograminearum alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Research 22: 4673–4680. van de Peer, Y. & De Wachter, R. (1994) TREECON for Windows: a software package for the construction and drawing of evolutionary trees for the Microsoft Windows environment. Computer Applications in the Biosciences 10 : 569–570.

420 Ward, T. J., Bielawski, J. P., Kistler, H. C., Sullivan, E. & O’Donnell, K. (2002) Ancestral polymorphism and adaptive evolution in the trichothecene mycotoxin gene cluster of phytopathogenic Fusarium. Proceedings of the National Academy of Sciences, USA 99: 9278–9283. Corresponding Editor: S. Takamatsu